Dangerous delay in responding to Horn of Africa early warnings of drought

Summary of briefing paper¹

According to a briefing paper just released by Oxfam and Save the Children UK (SC UK), the 2011 crisis in the Horn of Africa has been the most severe emergency of its kind this century, with more than 13 million people still affected and hundreds of thousands at risk of starvation. The authors assert that this crisis, affecting primarily the drylands of Ethiopia, Somalia and Kenya, unfolded despite having been predicted. All three countries were almost equally affected by the total failure of the October to December 2010 rains and the poor performance of the March to May 2011 rains, resulting in crop failure and animal deaths. The situation was much worse in southern and central Somalia, where conflict further impeded traditional drought coping mechanisms, and reduced access for humanitarian agencies. The briefing paper is principally concerned with how the international system responded (or not) to early warnings of the oncoming crisis, and why it was allowed to spiral into a disaster.

The countries affected by this drought were in very different situations. For example, in Ethiopia, there has been considerable effort to build resilience through the development of the Productive Safety Net Programme (PSNP), investment in new health posts which enabled huge increases in access to nutrition responses, investment in pastoral areas through the Pastoral Community Development Programme (PCDP) and the promotion of disaster risk management policy and practice. This contrasts with Somalia, where such work has been largely absent, due to access restrictions, a complex environment and the unwillingness of donors to invest.

The emergency in the Horn of Africa in 2011 was not a sudden-onset crisis. Forecasts of the impending crisis started in August 2010, as changing weather conditions linked to the La Niña phenomenon were confirmed. These predictions became more strident in early November 2010, when the October to December short rains were forecast to be poor. This prediction was accurate, prompting the Food Security and Nutrition Working Group for East Africa (FSNWG) to set up a La Niña task force. In December 2010, it stated that "preemptive action is needed to protect livelihoods and avoid later costly lifesaving emergency interventions" and called on the humanitarian community (donors, United Nations agencies, non-governmental organisations (NGOs)) "to be prepared now at country level."

Multi-agency scenario planning took place in February 2011. A Famine Early Warning Systems Network (FEWSNET) food security alert dated 15 March 2011 made it clear that the current situation was already alarming and would deteriorate further if the March to May rains were as poor as expected. It stated that even average rains would lead to a critical food security situation until May or June, and predicted "localised famine conditions [in southern Somalia], including significantly increased child mortality… if the worst case scenario assumptions are realised". The FSNWG also warned that "failure of the March to May rains is likely to result in a major crisis". At this stage, humanitarian actors were advised to begin large-scale contingency/response planning immediately and to implement expanded multi-sectoral programming. Yet this call was not adequately heeded.

The various early actions taken by national governments affected by the crisis are set out in the briefing paper, e.g. in Ethiopia, the government’s Agricultural Task Force, supported by the Food and Agriculture Organization of the United Nations (FAO), developed a roadmap for interventions in early 2011. Meanwhile in the regional autonomous state of Puntland, the President announced a drought emergency in November 2010, and called on the international community and aid agencies to provide humanitarian assistance. However, it is the international response which the briefing paper focuses mainly upon.

The donor response at scale was too slow. Figure 1 shows the level of humanitarian funding to Ethiopia, Kenya, and Somalia in the lead-up to the crisis. An increase can be seen after the first warnings in late 2010 and the UN Consolidated Appeals Process (CAP) appeal (November 2010). But it was only after major media coverage in June/July 2011, and after the UN declared a famine in Somalia, that donors drastically increased the funds available. Some donor representatives in the region were also aware of and acted on the impending crisis much earlier. For example, key donors in Kenya - European Commission Humanitarian Aid & Civil Protection (ECHO), UK Department for International Development (DFID) and United States Agency for International Development (USAID) - met in December 2010 to co-ordinate their initial response. Indeed, Oxfam accessed ECHO funding for work in Turkana, Kenya, in April 2011, and this was quickly scaled up in July. Nevertheless, most donors were not able to access funding at scale from their headquarters until malnutrition rates were at dangerous levels and media attention broke the story.

In Somalia, a further complicating factor was present: the international community failed to prioritise growing humanitarian concerns over political considerations, reflected in Oxfam and Save the Children struggling to find funding for work in South Central Somalia.

The UN’s humanitarian appeal in November 2010 seriously underestimated the number of people in need of emergency aid. This is partly because the timeline of UN appeals is not aligned with the seasons in the Horn of Africa. Assessments were carried out in September, before the failure of the short rains (which normally start in October) and did not take into account the future weather predictions. And for Somalia, more recent appeals were based more on what programming can be achieved (within the constraints of access and partners) rather than what funding would be required to avert disaster, thus potentially giving a misleading picture of needs within the country. The Consolidated Appeal (CAP) - a key document for marshalling donor resources - was only fully revised at the end of July 2011. This was clearly a factor in the failure to scale up the response early on. In Somalia, for example, the original 2011 CAP was set at $530m in late 2010. This was revised to more than $1bn by August 2011.

Many agencies, including Oxfam and Save the Children, had begun a small-scale response by December 2010, and tried to focus international attention on the impending crisis. But while some performed better than others, most agencies did not adapt their programming on a sufficient scale to meet the level of need over the following six months, and did not begin to respond at scale until after the 2011 rains failed in May. Some agencies declared the situation a corporate priority as early as February 2011, but this only happened in Oxfam and Save the Children at the end of June and early July 2011 respectively.

Pressing domestic, regional and international developments, including the conflict in Somalia, the Arab Spring uprisings, global recession, other crises such as the Japan earthquake/ tsunami, or donor fatigue, may have delayed the international community’s response to the drought.

While the early warnings were clear, the scale (numbers of people) and depth (severity) of the crisis still caught many by surprise. This is partly because needs assessments carried out by UN agencies or governments - which are a key driver for donor interventions - are published several months after the assessment is done and critically, do not incorporate forecasts or predictions based on a changing situation. Thus the UN appeal for Somalia, launched in November 2010, had relatively low figures for those in need of assistance in 2011 and failed to sufficiently reflect the La Niña predictions. Ultimately, the EWSs performed, but decision makers chose not to respond. Possible reasons for reluctance to respond early may include:

• Fear of getting it wrong - with both financial and reputational risk at stake.
• Fear of being too interventionist - undermining communities’ own capacities to cope.
• Fatigue - "there are droughts every year" - encouraging an attitude of resignation to the high levels of chronic malnutrition, and an inability to react to the crisis triggers.

All humanitarian actors - governments, UN agencies, donors, implementing NGOs - want to be certain about the scope and depth of a looming food crisis before responding at scale. The international humanitarian system only becomes fully operational when Integrated Phase Classification (IPC) phase 4 - ‘emergency’ - has been reached². Responding on the basis of forecasts instead of hard data requires a shift in dealing with uncertainty. Forecasts involve uncertainty. Yet this uncertainty is not unquantifiable - standard risk management techniques allow us to convert this uncertainty into risk, which can then be managed and minimised. Figure 2 shows a typical risk impact/probability chart, which plots the probability that a hazard will occur against its impact.

Using this logic, it would have been clear from around January 2011 that the high probability of poor March-May rains in the Horn of Africa, magnified by the failure of the previous rains in late 2010, would constitute a critical risk that needed to be addressed immediately.

Once the EWS has flagged a potential problem, this should immediately activate a process of further investigation - detailed monitoring which can be used to design interventions - and the operationalisation of emergency plans. These plans need to be clear on who should do what and when. There should be a common approach to using triggers, so that decision makers know exactly what they ought to be doing as the situation deteriorates and the consequences if they fail to act on those triggers.

All actors need to work together to develop a system of triggers that:

• recognises the national government (where possible) as primary duty bearer for meeting citizens’ food needs
• reflects the high levels of chronic malnutrition in some areas
• reflects the exponential rather than linear development of malnutrition
• does not lead to interventions that undermine communities’ capacity to
• cope
• is context-specific for different livelihoods zones
• is agreed between different actors

It is expected that there will be a range of triggers for different sorts of response. So, for example, at an early stage the trigger might be for advocacy, but as the situation deteriorates, it might be for a livelihood response, and subsequently for a food/nutrition response.

Long-term programmes are in the best position to respond to forecasts of a crisis. There are established links with communities and thus the vulnerabilities and the complexities are understood. There are staff and/or partners in place, they are in a trusted position with donors with funding available and their work has been negotiated with government bodies. Long-term programmes should become more sensitive to drought risks and seek to reduce vulnerability by reducing the underlying risk factors. When risk analysis is made integral to long-term programme design - by using Drought Cycle Management (DCM) or similar tools - droughts can be seen as an integral part of the livelihood system, rather than as an unexpected shock.

Early humanitarian response, which seeks to reduce disaster risk, is both effective and costeffective in addressing the underlying factors that make people vulnerable.

Humanitarian interventions which should be started on the basis of forecasts include:

• Livelihood protection interventions, in order to safeguard key assets.
• Interventions involving a significant time lag. If food distributions are likely to be necessary, the process of mobilising resources and arranging logistics should begin on the basis of early assessments, with quantities being revised at a later stage.
• ‘No regrets’ options: measures that build capacity and disaster preparedness but have no negative effect even if the worst forecasts are not realised (either because the cost is very low or because they will build resilience). This would include activities such as
  • putting human resources systems in place
  • talking to existing and potential donors and drafting outline proposals for response
  • building links with private sector partners, e.g. developing standing agreements with money transfer companies, starting tender processes
  • building and strengthening the capacity of local partners
  • starting to engage with all relevant stake holders, such as partners, local authorities and the women, men and children in communities
  • practical measures like assessing borehole operations, prepositioning stocks, market assessments, mapping the capacity and coverage of traders, etc
• Flexible funding measures that involve a level of expenditure commensurate with the probability of the crisis occurring. Projects could be designed and initiated with limited but sufficient start-up funding, with a commitment to more substantial resources as the crisis develops.

Early response requires us to move away from the traditional distinction within the aid system between development and humanitarian work. This approach, with different staff, mandates, skillsets, timescales, budgets and beneficiaries, is not valid in regions like the Horn of Africa. To increase the effectiveness of the aid system, this artificial gap must be bridged.

Skilled and experienced staff and partners are needed who are able to build risk analysis into their work and are thus able to adapt what they do, and how they do it, as the situation and needs change.

Much greater investment is needed in longterm joint efforts to strengthen government capacity, both in disaster risk management and coordination, but also in improving the ability of long-term development work in all sectors to build resilience.

Humanitarian and development strategies are often developed separately, whereas a risk management approach requires common thinking and planning. Where structures are institutionally divided, an effective coordination and integration approach with various mechanisms for direct cooperation, joint programming and implementation (in combination with shared learning cycles) can help to merge development and response.

Neither humanitarian nor development funding streams are ideally suited to the situations of chronic vulnerability that occur regularly in the Horn of Africa, where the situation is often in transition between humanitarian emergency and development. Humanitarian programmes are short-term, which doesn’t allow for longer-term planning, but they are usually fairly flexible in terms of programme approach and the ability to change expenditure. Development programmes are long-term but less flexible. Implementing agencies are required to predict their expenditure at the start, with often only a small contingency (for the EU, this is a maximum of five per cent). This is designed to boost accountability, but it does hinder flexibility and agile programming. Some emergency aid donors have made considerable efforts to be more flexible, with innovative funding mechanisms to support recovery and resilience. Other donors should emulate these models.

Donors - including the UN for the Central Emergency Response Fund - need to revisit their mandates and protocols for funding streams and continue to push the boundaries, so that they can disburse sufficient funds quickly to support early response.

The ‘bottom line’ of this analysis is plainly stated: predictions about the impact of the 2010-11 drought in the Horn of Africa were clear, and unfortunately, much of what has happened was preventable. The scale of death and suffering, and the financial cost, could have been reduced if early warning systems had triggered an earlier, bigger response.

¹A Dangerous Delay. The cost of late response to early warnings in the 2011 drought in the Horn of Africa. Oxfam, Save the Children UK. Joint agency briefing paper. 18 January 2012. Available from: http://policy-practice.oxfam.org.uk

²http://www.ipcinfo.org/